March 2017
Review of Current Conditions:
THE ECONOMIC OUTLOOK AND ITS IMPACT ON WORKERS COMPENSATION
The exhibits below are updated to reflect the current economic outlook for factors that typically impact workers
compensation. Each exhibit also provides some context for the outlook, relative to the historical data. Forecasts are
derived from Moody’s Analytics.
EMPLOYMENT GROWTH
Private employment growth slowed to 1.9% per year
in 2016 after increasing to 2.3% per year in 2015, the
fastest rate of growth since the recession. Even with
slower growth, over 2.2 million workers were added
to payrolls throughout 2016.
Education and healthcare, and professional and
business services showed the largest employment
gains. Trade, transportation, and utilities; and leisure
and hospitality also showed strong gains. Natural
resources and mining posted the largest decline
followed by manufacturing with a small decline. The
decline in manufacturing employment is of concern
for workers compensation because manufacturing
accounts for 15% of premium in NCCI states. See
below in Drilling Down for a survey of manufacturing
presence by state and factors impacting trends in
manufacturing output and employment.
Consistent with slowing employment growth, real
gross domestic product (GDP) growth also slowed
last year. Annual growth in real GDP for 2016 fell to
1.6% from 2.6% in 2015.
Private employment rose by 227,000 jobs in February 2017 following an increase of 221,000 jobs in January. Despite these
robust increases, Moody’s forecasts that employment growth will continue to slow to 1.8% this year and 1.6% next year.
The forecast for this year is two-tenths of a percentage point higher than in last December’s Quarterly Economics Briefing
(QEB). Moody’s has revised its forecast for 2018 from last quarter, when they expected growth to reaccelerate next year.
Growth in employment leads to increased premium, but inexperienced new hires may also put upward pressure on claim
frequency. Both impacts are likely to be muted over the next couple of years if employment growth slows.
Analysis and charts prepared in February and March 2017.
1
WAGE GROWTH
Wage growth is forecast to have slowed last year to
2.5% after increasing by 3.1% in both 2014 and
2015. This is an upward revision from the last QEB,
where we estimated average weekly wage growth
for 2016 at 2.2%. In addition, wage growth is
forecast to accelerate sharply to 3.9% this year and
5.3% next year. The forecast for 2017 is threetenths of a percentage point below that posted in
December.
Wages are forecast to accelerate due to tightening
labor market conditions. The unemployment rate
averaged 4.9% in 2016—down from 5.3% in 2015—
and fell further to 4.7% in February 2017.
The broader measure of unemployment that
includes discouraged workers and part-time
workers who would prefer a full-time schedule is
also declining. It averaged 9.6% in 2016 compared
to 10.5% in 2015, and declined to 9.2% in February,
the lowest rate of the recovery. While declining, it is
still high by historical standards.
Employers will likely need to offer higher wages to
attract workers as the unemployment rate declines. Higher wages will, in turn, lead to both increases in workers
compensation premiums and indemnity severity.
MEDICAL INFLATION
Medical inflation accelerated in 2016 to 3.8%, up
from 2.6% in 2015. This reflected an increase in
price inflation generally from 0.1% in 2015 to 1.3%
in 2016. Moody’s expects medical inflation to slow
slightly this year to 3.4% and continue at that pace
in 2018. This is still above the average of 3.0% for
the previous five years, and also higher than general
inflation in both years. Moody’s forecast for general
inflation is 2.8% this year and 2.5% next year.
Changes in medical severity are a function of
changes in both price and utilization. Higher medical
inflation (the price piece of that equation) would
imply upward pressure on medical severity.
However, in 2015, workers compensation medical
severity declined by an estimated 1% for NCCI
states 1 although medical inflation increased by
2.6%. A study in the 2016 Issues Report Fall Edition
States included are AK, AL, AR, AZ, CO, CT, DC, FL, GA, HI, IA, ID, IL, IN, KS, KY, LA, MD, ME, MO, MS, MT, NC, NE, NH, NM, NV, OK, OR, RI,
SC, SD, TN, UT, VA, VT, and WV.
1
Analysis and charts prepared in February and March 2017.
2
published on ncci.com found that a 3% decline in utilization of physician services was the major driver of the decline in
medical severity in 2015.
INTEREST RATES
Low interest rates have constrained investment
income in the property/casualty (P/C) industry for
many years. However, with price inflation returning
to more normal levels, interest rates are also
expected to rise.
Indeed, the Federal Open Market Committee
(FOMC) increased the target range of the federal
funds rate by 25 basis points to 0.75%–1.00% at its
March meeting after last increasing the target by
25 basis points in December 2016.
The chart on the right shows interest rates for the
10-year Treasury note as of June for 2012–2018.
This 10-year rate has been below 3% for the last
five years and declined to 1.6% in 2016 due to
uncertainty surrounding the British vote to exit the
European Union.
Moody’s expects the interest rate on the 10-year
Treasury note to increase to 2.7% in the second
quarter of this year and to 3.5% next year. The
expectation for 2017 is the same as shown in the
December QEB and is in line with Moody’s expected increase in inflation this year. Their expectation that the FOMC will
more aggressively raise short-term interest rates next year underlies their forecast for further acceleration in 2018.
The projected increase in investment yields will increase the potential contribution of investment income to total
profitability. However, medical inflation is also forecast to increase as shown in the previous chart, and may offset some of
the potential gains from investment income. The overall impact of increases in interest rates and medical inflation is
uncertain and will depend on the magnitude of each.
Analysis and charts prepared in February and March 2017.
3
DRILLING DOWN:
ECONOMIC TRENDS IN MANUFACTURING
In this edition, we focus on manufacturing. First, we show the importance of manufacturing to the workers compensation
industry. Next, we present a survey of state manufacturing presence in which we compare relative shares of manufacturing
employment across states. Manufacturing encompasses a wide variety of different industries, so we break down total
manufacturing into its subsectors focusing on differences in durable and nondurable industries. Analysis in this section
focuses on data for 2015 and 2016.
Finally, we look at long-term trends in manufacturing output, employment, and productivity back to 1990, and the rise of
automation and supply chain decentralization that has contributed to these trends. The supply chain discussion makes use
of the automotive industry as an example.
THE IMPORTANCE OF MANUFACTURING TO WORKERS COMPENSATION
Manufacturing is important to the workers
compensation industry because it makes up a
disproportionate share of premium relative to
exposure. In Policy Year 2016, manufacturing
made up just 8% of exposure in terms of
payroll, but 15% of total manual premium
across NCCI states.
As shown in Figure 1, manufacturing’s share of
total manual premium ranges from less than
10% for states with low shares of
manufacturing employment to more than 20%
for those with high or middle-high shares as
defined in the next section.
SURVEY OF STATE MANUFACTURING PRESENCE
In 2016, the US manufacturing sector employed more than 12.3 million workers. The first column in Table 1 shows the
number of manufacturing workers employed in each state. The second column shows the share of manufacturing
employment calculated by dividing manufacturing employment in a state by that state’s total private nonfarm employment.
Excluding DC, the average manufacturing employment share across states was 10.3% in 2016. (DC is excluded because it
has an immaterial manufacturing presence.)
Analysis and charts prepared in February and March 2017.
4
States are divided into four categories (high, middlehigh, middle-low, and low) based on their manufacturing
employment shares. States in the high category have
manufacturing employment shares that are greater than
one standard deviation above the average share across
states, and states in the middle-high category have
shares that fall between the average and one standard
deviation above the average. Middle-low and low
categories are defined similarly. Green indicates states
with above-average shares and red indicates states with
below-average shares. Darker shades denote states that
are more than one standard deviation away from the
average share across states; lighter shades indicate
states within one standard deviation of the average.
States with high manufacturing shares of total
employment are those above 14.5% while those with
low shares are below 6%. Note that even though
California employs the most manufacturing workers in
the country, its manufacturing employment share is
below average at 9.4%. On the other hand, smaller
states such as Arkansas and Mississippi employ fewer
manufacturing workers but have high manufacturing
employment shares of 15.3% and 15.9%, respectively.
The map in Figure 2 shows the geographical distribution
of manufacturing shares using the same categories and
colors as in Table 1. It indicates that states with high
manufacturing employment shares (dark green) are
primarily located in the Midwest and central South.
Many states with high manufacturing employment
shares tend to have high exposure to durable goods manufacturing, as we will see next. States with low manufacturing
employment shares (dark red) are distributed in both the West and East. In general, the economies in these states depend
more on services than manufacturing.
Manufacturing consists of many
different industries that can be
broadly grouped into durable goods
and nondurable goods. Durable goods
are defined as those that do not wear
out quickly and are used for an
extended period of time, usually
three or more years. Examples
include furniture, appliances, and
cars. Nondurable goods are ones that
are either consumed in a single use or
have a life span of fewer than three
years, such as food, apparel, and
paper products.
The key functional distinction
between the two categories is that
demand for durable goods is more of
an investment decision, and demand
Analysis and charts prepared in February and March 2017.
5
for nondurable goods is more of a consumption decision. Purchases of durable goods can be timed or postponed more than
purchases of nondurable goods in response to current economic conditions or future expectations. Therefore, demand for
durables can be expected to exhibit greater cyclical volatility than nondurables.
Because of this distinction, economic conditions will not affect all states with high manufacturing employment shares in the
same way. States with higher concentrations in durable goods industries can be harder hit by economic slowdowns. And
workers compensation premiums, frequency, and severity in those states could also be impacted more than states with
high concentrations in nondurable goods. For example, states with high concentrations in auto manufacturing may be
harder hit during a recession than states with high concentrations in food manufacturing.
The following tables are similar to Table 1 above showing employment shares, but break out the data for durable and
nondurable manufacturing for 2016. Table 2 is for the durable manufacturing sector while Table 3 is for nondurable. Again,
categories and colors are assigned based on the average share in each table and the standard deviation. The average
employment share across states is 6.4% for durable manufacturing and 4.0% for nondurable manufacturing.
Analysis and charts prepared in February and March 2017.
6
Maps for employment shares of durable and nondurable manufacturing are shown in Figures 3 and 4. Arkansas, South
Carolina, Kansas, North Carolina, and Nebraska are more important producers of nondurable goods than durable goods.
Therefore, the manufacturing sectors in these states should exhibit less cyclical sensitivity than Michigan, Mississippi,
Kentucky, and Ohio, which are more dependent on durable goods and less dependent on nondurables.
Table 4 drills down further to the three-digit North American Industry Classification System (NAICS) codes within durable
and nondurable goods. Data for this level of detail is from a different data source (the Quarterly Census of Employment and
Wages), for which 2015 is the latest year currently available. Employment shares are displayed for each manufacturing
subsector for the 13 states with high shares of durable and nondurable manufacturing employment from Figures 3 and 4.
Four of the 13 states have both high durable and nondurable shares.
Analysis and charts prepared in February and March 2017.
7
Colors are assigned based on a comparison of each state’s share to the US average for each three-digit NAICS code. Cells
shaded darker green indicate the manufacturing sectors which are most concentrated in each state. For example,
transportation equipment manufacturing in Michigan is dark green indicating it has a much higher share than the nation as
a whole. Kentucky, Indiana, and Alabama also have significantly higher shares in transportation equipment manufacturing.
Other industries with high shares relative to the nation for the eight states with high durable manufacturing shares include:
•
•
•
•
•
Wood product manufacturing—Mississippi and Alabama
Primary metal manufacturing—Indiana and Alabama
Machinery manufacturing—Michigan, Wisconsin, and Iowa
Electrical equipment and appliance manufacturing—Wisconsin
Furniture and related product manufacturing—Mississippi
For the nine states with high shares in nondurable manufacturing, dominant industries include:
•
•
•
•
•
Food manufacturing—Wisconsin, Iowa, Arkansas, Kansas, and Nebraska
Textile mills—Alabama, South Carolina, and North Carolina
Paper manufacturing—Wisconsin and Arkansas
Printing—Wisconsin
Plastics and rubber products manufacturing—Indiana, Wisconsin, Arkansas, and South Carolina
Analysis and charts prepared in February and March 2017.
8
TRENDS IN MANUFACTURING OUTPUT, EMPLOYMENT, AND PRODUCTIVITY
This section examines longer-term trends in US
manufacturing output, employment, and labor
productivity. Figure 5 contains cumulative changes
since 1990 in US manufacturing output and
employment. For decades, US manufacturing
output has grown while manufacturing
employment has declined. In 2016, manufacturing
output reached a record high. Since 1990,
manufacturing output grew 71.8% while
manufacturing employment fell 30.7%. This means
that in 2016 the United States produced almost
72% more goods than in 1990, but with only about
70% of the workers. A similar trend is seen since
2000 when output grew by 17.4% and employment
fell by 28.3%. In the post-recessionary period since
2010, manufacturing output grew 23.5% while
employment also grew but by only 7.0%.
AUTOMATION
Part of the divergence between output and
employment trends is due to increases in labor
productivity resulting mainly from capital
investment to automate production. Automation can include computerized machines, streamlined processes, and robots. 2
In 2014, there were 1.5 million robots in factories and warehouses worldwide, with the number expected to rise to 1.9
million this year. Japan has the most industrial robots at 306,000 followed by 237,000 in North America. 3
The long-standing trend that automation has decreased manufacturing costs over time has made US manufacturers less
expensive and more competitive, reducing the labor required to produce output. Figure 5 contains a line showing
cumulative changes since 1990 in US manufacturing labor productivity, or output per hour. Labor productivity has grown
140.1% since 1990, 62.5% since 2000, and 13.2% since 2010.
There are many reasons companies cite for automating manufacturing processes, with increasing labor productivity being
just one of them. Other cited reasons include reducing labor costs, lessening the impact of labor shortages, cutting out
routine manual tasks, improving worker safety, improving product quality, reducing lead time, and achieving competitive
advantage. 4 An article by BCG Perspectives states that in 2015 only about 10% of manufacturing tasks globally were
performed by robots, but that is expected to rise to 25% by 2025 as robots become less expensive and easier to program,
making them more accessible, particularly to small factories. Four manufacturing sectors are expected to account for 75%
of the expansion, including computers and electronic products; electrical equipment, appliances, and components;
transportation equipment; and machinery. 5 In fact, automation is starting to take hold in other economic sectors beyond
manufacturing, including kiosks and tablets to place orders and pay in restaurants, robots to process packages in
warehouses, and self-driving trucks in transportation. 6
Newman, Rick, “How Robots Paved the Way for Donald Trump,” Yahoo Finance, July 14, 2016.
West, Darrell M., “How Technology is Changing Manufacturing,” Brookings, June 2, 2016.
4 Csanyi, Edvard, “9 Reasons for Automation of Manufacturing Processes,” Electrical Engineering Portal, January 11, 2016.
5 Sirkin, Harold; Zinser, Michael; and Rose, Justin; “The Robotics Revolution: The Next Great Leap in Manufacturing,” BCG Perspectives,
September 23, 2015.
6 Newman, Rick, “How Robots Paved the Way for Donald Trump,” Yahoo Finance, July 14, 2016.
2
3
Analysis and charts prepared in February and March 2017.
9
Automation has reduced the cost of US manufactured goods, making manufacturers more competitive, but it is also the
main driver of decreased employment in manufacturing. A Ball State University study found that 87% of the job losses in
manufacturing from 2000 to 2010 are due to automation that led to productivity growth, and 13% are due to globalization
and trade. 7 Increased automation also means that today’s manufacturing jobs require skills that go beyond the old-time
blue collar jobs of the 1970s. More than half of manufacturing workers now have some education past high school versus
only a quarter in 1979. 8
DECENTRALIZED SUPPLY CHAINS
Besides automation, a second trend has been the development of decentralized supply chains. Manufacturers today
achieve lower costs by producing and assembling different components in different locations, but this trend has also
contributed to a decline in US manufacturing employment to the extent that production is located outside the United
States. The North American Free Trade Agreement (NAFTA), effective since 1994, has allowed domestic auto manufacturers
to reduce costs by producing in neighboring countries rather than overseas, leading to a supply chain that spans the United
States, Canada, and Mexico. The US automotive industry provides an example of decentralization in the modern
manufacturing supply chain.
According to a January 2017 report by the Center for Automotive Research (CAR), 9 the US automotive industry imported
$44.3 billion of auto parts produced in Mexico and $14.9 billion of auto parts produced in Canada during 2015, the latest
year for which data are available. The table below shows the percentage of parts coming from Mexico and Canada as a
proportion of all imports for different categories of auto parts.
Type of Auto Part
Share of Imports From Mexico and Canada
Engine and engine parts
59.8%
Transmission and powertrain parts
47.0%
Electrical and electronic equipment
60.7%
Steering and suspension parts
52.8%
Seating and interior trim
76.0%
Brake systems
36.9%
Automotive lighting equipment
37.8%
Motor vehicle parts not elsewhere specified
55.9%
Auto parts may cross borders as many as eight times before being installed in a final assembly plant in one of the three
countries. In fact, US content makes up 40% of the value of light vehicle and parts imports from Mexico and 25% of the
value from Canada. Overall, 51% of the value of US auto parts imports is from Canada and Mexico, with three-quarters
coming from Mexico and one-quarter coming from Canada. These statistics indicate that parts sourced from Mexico and
Canada are integral to the US auto industry.
Hicks, Michael J. and Devaraj, Srikant, “The Myth and the Reality of Manufacturing in America,” Ball State University, June 2015, p. 6.
McGill, Andrew, “The Impossibility of Reviving American Manufacturing,” The Atlantic, April 28, 2016.
9 “NAFTA Briefing: Trade Benefits to the Automotive Industry and Potential Consequences of Withdrawal from the Agreement,” Center
for Automotive Research (CAR), January 2017, www.cargroup.org/?module=Publications&event=View&pubID=148.
7
8
Analysis and charts prepared in February and March 2017.
10
The map in Figure 6 shows the interconnection of automotive supply chains. There are approximately 6,500 original
equipment manufacturer and supplier locations across the United States, Mexico, and Canada. The squares are original
equipment manufacturers (OEM) and can be both automaker assembly plants and automaker parts plants. Blue indicates
the plant is one of the former three Detroit automakers (Ford, GM, or Chrysler [now Fiat Chrysler]) and green indicates all
other automakers. The red circles indicate automotive suppliers.
The map illustrates that automotive parts and components are produced in all three countries. These decentralized supply
chains have allowed automotive manufacturers to locate assembly and parts plants in the location with the lowest costs,
increasing their competitiveness, but also contributing to the decline in employment shown in Figure 5. Because of NAFTA,
US automakers have structured their US operations interdependently with Canadian and Mexican suppliers, some of whom
are OEMs themselves.
Analysis and charts prepared in February and March 2017.
11
KEY TAKEAWAYS
In this report, we have focused on the manufacturing sector. Key takeaways include:
•
•
•
•
•
•
Manufacturing is important to the workers compensation industry because it makes up a disproportionate share of
premium (15%) relative to exposure (8%).
States with high manufacturing employment shares are primarily located in the Midwest and central South. States in
the East and West tend to be more service-oriented, with lower shares of manufacturing employment.
Durable manufacturing is more cyclical than nondurable manufacturing. Therefore, states with higher concentrations in
durable goods industries can be harder hit by economic slowdowns.
For several decades, US manufacturing real output has increased while employment has declined. In 2016 the US
produced almost 72% more goods than in 1990, but with only about 70% of the workers.
Increases in automation have reduced manufacturing costs, making US manufacturers less expensive and more
competitive, and reducing employment required to produce output.
Decentralized supply chains allow manufacturers to achieve lower costs, but also contribute to a decline in US
manufacturing employment to the extent that production is located outside the United States.
© Copyright 2017 National Council on Compensation Insurance, Inc. All Rights Reserved.
THE RESEARCH ARTICLES AND CONTENT DISTRIBUTED BY NCCI ARE PROVIDED FOR GENERAL INFORMATIONAL PURPOSES ONLY AND ARE PROVIDED “AS
IS.” NCCI DOES NOT GUARANTEE THEIR ACCURACY OR COMPLETENESS NOR DOES NCCI ASSUME ANY LIABILITY THAT MAY RESULT IN YOUR RELIANCE
UPON SUCH INFORMATION. NCCI EXPRESSLY DISCLAIMS ANY AND ALL WARRANTIES OF ANY KIND INCLUDING ALL EXPRESS, STATUTORY AND IMPLIED
WARRANTIES INCLUDING THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
Analysis and charts prepared in February and March 2017.
12